Real Time Automatic Detection of Motorcyclists With and Without a Safety Helmet
Valanukonda Lakshmi Padmini, G. Krishna Kishore, Ponnuru Durgamalleswarao, Parasa Teja Sree
Abstract
In the developing countries like India, the motorcycle riders are increasing day-by-day, wherein it also constitutes to the unprecedented increase in the number of motorcycle accidents across the country. To overcome this drawback, the proposed research work explains and demonstrates a method to enforce better safety protocols through the automatic detection of motorcyclists with and without a safety helmet by using a real-time traffic surveillance footage. The real-time automatic detection of motorcyclists with and without a safety helmet is established through detecting a vehicle and track pipelining it with OpenCV, sklearn, utilizing a descriptor known as the histogram of oriented gradients (HOG), and support vector classification (SVC), which are the combination of tools pertaining to machine learning and image processing mechanisms. With OpenCV Library method, a bike rider is identified in the surveillance video. Further by using a popular machine learning algorithm model called LinearSVC, the classifier label identifies whether the rider is wearing a safety helmet. The data attained in correspondence to the count of bike riders with and without safety helmet is stored in MySQL database with respective timestamps and is also visualized through tabular and graphical views in the developed desktop interface application. With 87.6% model accuracy, our paper proposes a solution to enhance the existing safety measures and provide a time-efficient approach to handle traffic regulations.